@declan90 do you know if the quoted rad BTU/KWH output is at a specific flow temperature?

having a combi boiler we only heat the water we use…when the heating is off in the summer, and subtracting the standing charge, do you think it’s reasonable to extend those costs across the year for water heating, and take that off the 13600 they think I’ll use for space heating?

@sicky This one is not too bad if you have all the correct data. It is just tedious getting the data and filling it all in as there is a lot of detail. You need all the u-values for non-standard elements. However there are drop-down options you can select for most cells e.g. cavity/solid wall 9", single/double glazed PVC/wood frames, suspended wood floor, floor insulation, etc. You must have all dimensions of rooms, windows, doors, etc. It calculates the oversized radiators and pipe spacing you need. It is professional and very good. Look at the picture of summary sheet I posted. If you look and want to have a go, I can send you a copy of the main sheet filled in.

Posted by: @sicky

↑

interpreting what it’s actually telling you and what to do with that information is where I feel like I’m a bit lost

The empirical method is actually quite simple, and it gives you a clear answer, as long as you use data from a period when the indoor air temp (IAT) stayed stable. You only need OAT (outside air temp, for the x axis) and energy out ie energy delivered to the building for the Y axis and the you do a scatter plot, and then add a trend line. Where the trend line crosses your Y axis at your design OAT on the X axis is your design design temp. I will dig out one of my charts tomorrow and post it.  

Posted by: @sicky

↑

@jamespa what size/depth insulated board on the walls/ceiling? Did you also add loft insulation on top?

 

I have done various thing, basically my approach was 'do a sensible and practical upgrade whenever a room is decorated or the opportunity arises'.  Over a period of 15 years this has transformed a cold house into a cosy one.  The starting point was a solid (9") wall 1930s house with a couple of breeze block extensions.

• Typically I used about 50mm celotex (plaster backed so total thickness ~62mm) internal wall insulation so that I didn't lose too much room space.  It is said that you need to be careful not to create a void in which vapour is likely to condense - I haven't encountered a problem but I was fairly careful to seal.  Read the product instructions; some products come with vapour barriers.  Many should be screwed and glued, glue to seal and hold tight, screw so they dont fall off in a fire exposing the highly flammable insulation.  I still have a couple of lengths of uninsulated wall in the house, but nothing much to speak of.  if you have space for thicker wall insulation, up to say 100mm, consider fitting it.
• Between utility room (a garage conversion) and garage there was space for 100mm celotex, on the cold (garage) side, so I did that.  I backed it with plywood on the garage side which makes for good wall storage.
• My accessible loft area has been upgraded to 300mm rockwool
• A flat roof where the felt needed replacing was upgraded from a 'cold' construction to a 'warm' construction with 100mm celotex on top of the joists
• Another flat roof that I am unlikely to replace for a decade or more was upgraded internally with 50mm celotex, screwed to the internal ceiling
• I replaced several double glazed units (glass only) which were blown, switching them for low e argon filled.  The difference was so noticeable I did the rest.  I didn't change the (1980s uPVC) frames.  This made a much larger difference than ever I imagined it would.
• Three 'feature' metal framed windows have been retained for their visuals, but secondary glazed with magnetic fit perspex (which has a low visual impact)

The mistake I made, in retrospect, was not to insulate a concrete floor that was dug up for another reason.  Even 10mm would have made a difference - silly me.  Im not sure there is that much heat loss but the floor itself feels cold, and its in a room where that matters.

I haven't attempted to insulate the suspended floors.

Its probably not surprising, given this mish-mash, that the fabric spreadsheet calculations for the ASHP were utter nonsense, although I was careful to tell surveyors what the construction was (simplifying the description so they had a chance to note it - which they clearly didn't).

Insulation costs a lot and can be disruptive, but the disruption can be minimised with sensible timing and cost can be spread if that helps.  The gains in comfort are tangible and of course its a substantial saving in running cost.

Obviously a full external wall insulation would have been technically better but it didn't suit us and Im happy with where we got to.

Hope that helps.  

Here is the measured data (2 years worth of half hourly meter data) my selection was based on.  You don't need 2 years worth and the 24, 12, 6,3 hour plots are optional, but give added confidence.  I plotted against degree days because, at the time, I couldn't find a way to download average OAT free of charge.  I had a 15 year old boiler running in condensing mode most of the time, so the efficiency might be as much as 110%.  Since the consumption includes DHW I didn't see the need to add anything for that.

Here is my data for the last heating season, a simple plot of hourly OAT against energy out (delivered to the house), for all hours in days without a setback (setbacks alter the dynamics, for a start the IAT isn't constant), heating hours only ie DHW hours excluded, because they often have a small amount of space heating which distorts the plot: 

The energy out when the OAT is -2 (my design OAT) is about 7.75kWh (or put another way, using kW not kWh, 7.75kW for one hour).

However, the left hand side of the plot is a bit squiffy, with the points tending to sag somewhat below the linear regression line. This can be seen more clearly in the residuals (actual energy out - predicted energy out) for each OAT point:

Note how on the left hand side the residuals ten to be negative, ie the actual energy in was less than the predicted energy in. This happens for a variety of reasons, including defrost cycles, and, in my system, the heat pump is at its limits, and struggling, most likely because the primary pipework resistance is too high. To remove this skewing effect, I can plot only the hours when the OAT was 5 degrees and above:

The correlation stays pretty much the same, but the left hand end of the regression line moves up a bit, and the heat loss at -2 degrees OAT becomes around 8.6kWh (or 8.6kW for one hour). This is my empirically derived heat loss (to which I have to add a bit for DHW heating. It is a lot less than the Freedom spreadsheet derived figure of 12.3kW (and as it happens bring my heat loss down below what my heat pump should be able to put out at low OATs).  

Note that the first chart is 100% based on observed real world data. There is no whatiffery involved whatsoever. Even fitting a regression line (the language we used in my day) is a descriptive exercise, the equation describes an observed relationship, not a modelling exercise, even if in the modern slack lingo folks may talk about 'modelling' the energy out against the OAT. The third chart does use a small bit of modelling/whatiffery ('what if the OAT was - 2 degrees?') but it is all very transparent, and no one is asked to stretch their creditabilities beyond belief, all the more so given what can be seen in the first chart.      

I happen to have a lot of data, because I have been monitoring my system for some time. But even a much smaller data set will give you a good idea of what your heat loss is. It is certainly worth doing if you can get the data.  

I’ve used a few and in my estimation:

* heatpunk.co.uk - nice and visual and good usability. 
* carno.io - no floor plan but otherwise nice and easy to use. I would say on a par with heat punk.  
* heat-engineer.com - not free and somewhat difficult to use, lots of forms to fill in and not always obvious to me what needs to be filled in. 
* build your own spreadsheet - good for learning how u-values, heat loss etc works, horrible in every other way. 

overall I’d say it’s not that hard if you have done any floor planning before and have some knowledge of building material. One thing to note is that the air exchanges are usually set too high (1.5-2) where a realistic value is 0.5-1, which in a reasonable to well insulated but old house leads to an overestimate, so you need to adjust that. 

Posted by: @adrian

↑

One thing to note is that the air exchanges are usually set too high (1.5-2) where a realistic value is 0.5-1, which in a reasonable to well insulated but old house leads to an overestimate, so you need to adjust that

I completely agree based on my personal experience, and I have now heard a sprinkling of installers say this, but is there any wider evidence that you know of.  This can be a really important factor in the calculation and yet it seems, based so far only on sparse evidence, that it may be systematically overestimated.

Posted by: @jamespa

↑

it may be systematically overestimated

I think it is possible installers may consciously or subconsciously welcome this, because it allows them to over-size on the back of what looks like a jolly scientific approach to determining heat loss. Most potential heat pump buyers won't spot the fact that the calculations are nothing more or less than classic whatiffery, which by definition allows a lot of what-iffing. I know of one installer who did an accidental what if, what if the kitchen external wall was 2m long rather than 12m long (it was a typo). Another installer's heat loss calculations amounted to not much more than an exercise in random number generation, which at least could be spotted because the numbers were so obviously wrong. But as I have said before, the potential buyer who approaches the heat loss calculation sceptically is likely to be relatively rare, or to put it another way, most potential buyers are at the mercy of whatiffery, or those five most dread words in the English language, 'leave it to the professionals'.         

I agree, it is weird that we are suppose to trust this whatiffery, but not our own usage data. 
after all, as was stated above, assuming a linear relationship between inside-outside temperature difference and consumption, it is very easy to calculate the heat loss at design temp

heat loss=average kWh usage*efficiency*(21-design temp)/(21-average outside)

Posted by: @jamespa

↑

is there any wider evidence that you know of. 

just some comments by a door blower tester that they have never seen a higher value than 1 and the fact that in my calculations I could only get close to the measurement value by setting the air exchange lower. Also I’d argue that if you know the fabric of the house, the rest of the calculation should be correct, apart from secondary heat gains (humans, sunshine, cooking etc). 

Posted by: @adrian

↑

I agree, it is weird that we are suppose to trust this whatiffery, but not our own usage data. 
after all, as was stated above, assuming a linear relationship between inside-outside temperature difference and consumption, it is very easy to calculate the heat loss at design temp

heat loss=average kWh usage*efficiency*(21-design temp)/(21-average outside)

Based on my own experience there are installers out there who will use more realistic values for ACH and/or take note of measured data.  They are in the minority for certain, but perhaps a growing minority.  My advice to anyone who has reason to believe that the GIGO fabric spreadsheet is giving a wrong answer is definitely to find another installer (which is what I did).

I challenged MCS on the matter a year or so ago.  The response was interesting, they admitted that there are problems with the GIGO fabric spreadsheet method, but said that at the present time they don't have a recognised alternative.  The GIGO fabric spreadsheet is based on a British Standard designed, as I understand it, for calculating new build.  The vast majority of the heat pumps we need to install are retrofits so that alone should ring alarm bells.  However as @cathoderay says, many installers probably welcome built in oversizing, and MCS can continue to 'hide' behind the BS (interpret that how you will) to defend their position.

Not a pretty position but one that, I fear, is not likely to change for some while yet.  Octopus and BG, who of course have access to smart meter data if you take your energy from them, may just burst through at some point, driven by the commercial reality that selling almost everyone (say) either a 8kW or a 4kW heat pump would make the whole thing extremely simple.  I'm firmly convinced that there must be a rich seam of information in smart meter data which, given a bit of intelligence (artificial or human) could be mined to find out rather more about the thermal properties of houses that we currently understand.